How We Evaluate Facial Coverage

Why coverage needs a method

Coverage sounds simple until you try to measure it. Two masks can both claim “full face coverage” and treat noticeably different areas. A manufacturer diagram showing a face bathed in even red light is an illustration, not a measurement.

Coverage also cannot be separated from fit. A mask that reaches the jawline on the face it was designed for may not reach it on yours, and a mask that sits further from the skin in one region is delivering less light there even where it nominally covers. So the method has to look at the physical unit, on more than one face.

The zones

We assess against a fixed list of regions, the same for every mask:

1. Forehead — including the upper forehead near the hairline, which rigid shells frequently leave short.
2. Periorbital (around the eyes) — treated or shielded, and which.
3. Nose — the bridge and the sides, which behave differently in rigid designs.
4. Cheeks — the largest single area, and the one most masks handle well.
5. Perioral (around the mouth) — commonly under-covered.
6. Chin.
7. Jawline and the sides of the face — where flexible and rigid constructions diverge most.
8. Neck — only where the device claims it or ships an attachment.

Each zone is recorded as fully covered, partially covered, or not covered, along with a note on why.

What we do with the physical unit

We photograph the mask in position on more than one face, from the front and in profile, and publish those photographs with the review. A reader should be able to see the gaps rather than take our word for them.

We record the emitter layout. LED spacing and distribution across the surface, including whether any region has visibly sparser coverage than the rest. An area with a wide gap between emitters is not receiving the same intensity as one directly in front of a dense cluster.

We note where the mask holds LEDs away from the skin. Standoff distance varies across a rigid shell, and where we can establish it, the review says so. This is the mechanism by which two masks with identical claimed irradiance deliver different amounts of light to your cheekbones — see how to read irradiance and dose claims.

We check fit across face sizes. A mask assessed on one face tells you about one face. Where a device performs noticeably differently on a smaller or larger face, that goes in the review, because it changes who the mask suits.

We record the strap adjustment range, since a strap that cannot be tightened enough lets the mask sit further from the skin than the designer intended.

What we do not do

We do not accept manufacturer coverage diagrams as evidence. They are illustrations produced by the seller.

We do not infer coverage from LED count. A high count concentrated in the centre of a mask is not broad coverage.

We do not present a coverage percentage. A single percentage figure implies a precision we cannot support and hides the thing that matters, which is which regions are covered. A mask covering 80% of the face while missing the entire jawline is not equivalent to one covering 80% distributed evenly, and one number would make them look identical.

We do not score coverage before we have held the device. Until then the field is empty, which is why it is empty across this whole site today.

How it feeds the score

Coverage carries 15 of the 100 points, weighted equally with fit and comfort, and second only to light output and specification transparency at 20. Together, coverage and fit account for nearly a third of a mask’s total score, which reflects a simple view: a specification that does not reach your skin is not a specification you bought.

The full scoring methodology sets out how the category is scored 0-10 and how that becomes a weighted contribution.

Reading a coverage assessment

When our reviews start carrying coverage findings, they will read like this: a zone-by-zone table, a photograph of the mask in position, a note on standoff distance where we could establish it, and a plain statement of which face shapes the mask suited and which it did not.

If a coverage assessment on this site ever consists of a percentage and an adjective, we have stopped doing the work described on this page, and you should tell us so via our corrections policy.